Initialize All Data.

Be sure you have backed up everything if you want to keep Library or User bank data.

WHY?

Explain WHY!!! that way the reader has a dozen more ways and chances to remember.

There is new content that doesn't land except for the initialize step. The act of initialization pushes certain content forward into different memory areas that the firmware update step alone doesn't do. That said - if the first thing you do is restore your previous configuration using an X7A (or X8A for MODX) backup restore - then I think all of this new content would be blown away anyways.

I personally skip the initialize all step knowing that there is some content I don't get to see. Some new Performances, for example, may depend on pre-loaded sequences that won't be there unless following the initialization step (and then immediately going and previewing that data).

Probably the "best" path would be to follow the instructions as-is but hold off on restoring your backup. First, go through the Performances inside the preset Live Set for the firmware update. BTW: this is where skipping versions you may lose out on documentation through the incremental firmware update Live Set "see what we did this version" stuff. And it's these Performances that tend to have the content that would be available through the initialize step.

More background can be gathered here:

https://www.yamahasynth.com/ask-a-question/just-updated-modx-to-2-5-with-one-question

I can't answer the philosophical question regarding why one would have to go through this kind of trouble to have content available to them. That's a battle to take up with Yamaha in order to try to get them to think more like an end-user and less like a software developer. Just because there's a bridge to a feature doesn't mean the trip to get there is a reasonable one. It seems actually what is sometimes built is a way to generate more work for the support team to field why certain presets don't work. Or something like this where it's not clear (because it wasn't documented) why one would perform the initialization step or, as I do, opt out of this step.

... and (no guide as to) why - after the initialization step (if this is what you do) - you would have a window of finding the data (with no real roadmap to find except for perhaps ... perhaps the Live Set presets) before you possibly trample over this by loading a backup. I don't see instructions that guide a user to take the clean initialized system and look at "this" and "this" before trampling.

A couple of general guidelines:

1) Always expect that any given firmware update will erase all of your data. So always make a backup unless you have nothing to save. Some updates will not erase data - but you do not want to rely on luck.

2) To experience all of the new content, initialize all data after a firmware upgrade. This is an optional step only to be sure that example patterns or other content not automatically landed by a firmware update is made available. I would say you have a window of opportunity to utilize this content before restoring your backup (X?A file). You do not have to do this initialization step if not concerned about some of the presets depending on these patterns breaking. I don't like how this part of the update process works - but that's how it works. The amount of content that breaks is fairly limited.

3) The Preset live sets are probably going to be the best bet for finding new content for the updated firmware that would depend on the initialization step. If anything in the preset live set doesn't seem to demo correctly - then chances are this live set Performance depended on an initialization that wasn't done.

Your 2nd point, above, grates.

It's ridiculous that an update's primary demonstrative content is "hidden" in this manner (firstly) and then destroyed by the user when re-installing their own prior works.

This is the kind of dark pattern that makes one wonder if anyone at Yamaha responsible for these products actually uses them.

The "file system" has always been a wonky thing to use. From what I see it's because of the lack of abstraction of how the underlying memory technology works. There is very little code put between the memory - and how that fabric itself organizes itself - and the GUI on top to manage this. So we end up with extra steps and strange "gotchas" that seem a bit unreasonable for the modern musician. There was a time when one had to almost co-design the hardware with the manufacturer. Because there was little documentation available to anyone and the usage hadn't been quite all mapped out yet. But today - several generations into the same technology with very little new (Superknob has been done before, assignable parameters have been done before, FM has been done before, AWM2, etc... MSeq is an extension of other past ideas not completely new) - the File System should have evolved to something more elegant. Since it's fairly stuck - it's also familiar. Those who have used previous products see some sameness of the hoops one has to jump through to do something that the rest of the world abstracts and makes simple.

I see this as a huge opportunity for Yamaha. That's the nice way of saying how bad I see this part of the design is. There's kind of two faces here. You see one philosophy of thinking like a musician and let the instrument speak to you in some organic way. However, there are these strange funky interruptions to that trance where you have to wake up and do calculus - something kind of not-so-musical. Every year technology makes it easier to hide complexity of hardware. Our sons and daughters can use a computer (the HD TV) with a database (Netflix) and pull up their cartoon programs years before starting grade school. Technology on so many fronts is getting out of the way but some gear is slow to adopt trends.

I really think it's too early (now this is me talking for myself) - to expect the ship to steer in the opposite direction. I see some improvements to other areas. The filesystem is still as screwy as ever. But other parts are starting to get better. I'm optimistic that future file systems will be more elegant. How one moves saved units of memory (patches, Performances, etc) to different areas. How updates work. All that stuff. Someday I imagine a time when this will be better. Sometimes the pendulum swings too far with abstraction. I don't want a smart device that prevents me from doing something that can be done now because it thinks I don't need that. I personally do not see elegance and power as mutually exclusive.

At any rate - I stay positive. I see the shortcomings and believe there are others inside Yamaha that see the same things. I believe there are those working to undermine some of the legacy that doesn't make sense anymore. Not because I have any inside information. That's part of the optimism. I also don't see any clear winners out there. You need to decide which fatal flaws you can live with. For now, I'm choosing Yamaha. Next year may be interesting for some of the manufacturers who haven't refreshed in a while.

For the record - I don't know for a fact that loading the backup file gets rid of patterns brought to the forefront by an initialization. This is my assumption given that it's supposed to replace "all" (except the firmware itself).

"For the record - I don't know for a fact that loading the backup file gets rid of patterns brought to the forefront by an initialization. This is my assumption given that it's supposed to replace "all" (except the firmware itself)."

Unfortunately it does. Some of the demo patterns are quite informative, too. Which is a great pity. And they're much more transparent than arpeggios.

It's impossible to do a nice merge, of all the things from a "user backup" without overwriting a tonne of stuff, not just patterns.

The Motif XF allowed for loading backups without waveforms. The concept of loading less than "all" from an all file has solved issues in the past. I'm not sure why this general train of thought wasn't carried forward. That's not as elegant as a "merge" - but choosing not to load say Patterns would at least give an option closer to this.

At any rate - in keeping with the system as it stands today - you would need to have backups of all of you Libraries and keep track of which one to load into which slot. Then restore everything back the more manual way of reloading Libraries and your user file individually. I'm not sure if restoring a user file blows away the patterns (I don't think it should) - but if it does then you'll need to save your user bank as a Library and restore this Library first, then import it to the User area (which is a merge event) delete the Library, then go about restoring your real Libraries in the correct order. If you have songs you want to keep - you'll have to manage those too.

This is one of those bridges that is possible to cross - you can get your old data back with the patterns in-tact - but the squeeze isn't worth the juice as they say.

One can come up with a number of theoretically better schemes to help manage this situation. I had one and erased it because it doesn't necessarily move the needle here.

On a tangent ...

I wish the reference manual was updated with supplementary manual releases because the supplementary manuals are incomplete. The reference manual has a chart for what is saved is in each file (X8L, X8U, etc) but supplementary manual for patterns doesn't update this graphic.

This is why it would be good if Montage Connect could save and restore Patterns the way it can currently do for Performances. Maybe in X7P files, analogous to the current X7B files for Performances.

I posted this thought to IdeaScale. Hopefully someone who can do something about it will see it and take action.

Jason is correct. The way recent features have been implemented shows clear evidence of the software folks tacking on capabilities in kind of a Rube Goldberg way, instead of doing what they should do and reworking the underlying “file system“ so that it can handle the new features in a more straightforward way.

I understand that hackery can get new features of the door more quickly, but there’s a price to pay that eventually gets really high.